This study's hypotheses are that rCD4-IgG will be safe and tolerable when administered to HIV-infected pregnant women and their newborn infants, and that an acceptable dosing schedule can be determined based on pharmacokinetic studies.